A Theoretical Review on Improving Predictive Accuracy and Mitigating Overfitting in Materials Informatics

The importance of data standardization from the viewpoint of various enterprises is explored in this study. In this article, we'll look at how data standards have evolved over time and how application programming interface (APIs) have become the de facto norm. Promoted system-to-system interoperability, less translation hurdles, and elimination of missing data issues are just a few of the benefits of data standardization highlighted in the paper. Multiple approaches to data normalization are investigated in the research as well. These include maximum scores, logarithms, and z-scores. This study compares three different standardization approaches to cluster analysis, looking at their effects and weighing the pros and downsides of each. Also included are the results of tests conducted with two databases and a study of data standardization as it pertains to Marajoara ceramics. The findings show that certain standardization methods work well when looking for correlations between different variables. Data standardization and its implications in many academic and corporate contexts are thoroughly examined in this work.